Graft versus host disease (GVHD) is the leading cause of morbidity and mortality following allogeneic hematopoietic stem cell (HSC) transplantation, an established therapy for patients with hematological malignancies. Current strategies to diminish GVHD include T cell depletion and immunosuppressive drugs, which are associated with an increased risk of tumor relapse, opportunistic infection, and/or toxicity. Novel approaches acting intrinsically on the immune system are clearly needed. In this regard, naturally occurring immunosuppressive cells, myeloid-derived suppressor cells (MDSCs), have recently gained considerable attention. MDSCs suppress T-cell responses through multiple mechanisms, e.g. iNOS, arginase, and reactive oxygen species. We demonstrated that MDSC induced tumor antigen specific T regulatory cells and T-cell anergy in vivo. Our preliminary results indicate that MDSCs have several attractive attributes as helper cells to inhibit GVHD without significantly compromising graft-versus-leukemia/lymphoma (GVL) in a murine model that results in the establishment of chimerism and long-term survival. The preliminary study showed that a significant number of MDSCs could be mobilized and expanded in the periphery. The objective of this proposal is to an optimized protocol by which MDSCs can be mobilized and expanded and to test the applicability of mobilized MDSCs in suppressing the allo-immune response without significantly suppressing the desirable GVL activity. Based on the results of our preliminary studies, we hypothesize that: (i) A significant amount of MDSC, exhibiting comparable suppressive functions as the tumor-host-derived counterpart, can be mobilized from bone marrow; (ii) GA can modulate differentiation and suppressive function of MDSC, thereby enhancing the efficacy of MDSC treatment in preventing GVHD; (iii) MDSCs derived from mobilization protocols can strongly suppress allo-responses mediated by CD4 T cells and induce Treg expansion, but exhibit less suppressive effect on CD8 T cells; (iv) MDSC treatment preferentially eliminates primarily activated T cells and skews toward the selective expansion of CD44+CD62L- memory CD8 and CD4 T cells, thereby preventing GVHD without significantly compromising the GVL activity; (v) RAE-1 (NKG2D ligand) expression by tumor cell is enhanced upon irradiation, which leads to increased susceptibility to killing by NKG2D+ CD8 T cells and/or NK cells. Three specific aims will be pursued: 1) Mobilize and expand myeloid-derived suppressor cells from bone marrow into the periphery and assess the prevention of GVHD by mobilized MDSCs in combination with GA treatment; 2) Study the mechanisms underlying the inhibition of GVHD mediated by mobilized MDSC; 3) Study the mechanisms underpinning the preferential suppression of GVHD by MDSC without significantly compromising GVL activity in pre-existing tumor models. The information will provide the basis and scientific principles for the mobilization and expansion of MDSC in human that can be used in clinical settings for future clinical trials.